The Political In‡uence of Peer Groups:
Experimental Evidence in the Classroom
Camila Camposy
Shaun Hargreaves Heapz
Fernanda L L de Leonx
November 23, 2013
Abstract
People who belong to the same group often behave alike. Is this because people with similar preferences naturally associate with each other or because group
dynamics cause individual preferences and/or the information that they have
to converge? This is the question that we address with a natural experiment.
We focus on the possible in‡uence of peers on two types of individual political
behaviour: political identi…cation on a left-right spectrum and political engagement. We …nd no evidence that peer political identi…cation a¤ects individual
identi…cation. But we do …nd that peer engagement a¤ects individual engagement, individual political knowledge and political identi…cation among those who
are initially least engaged. We argue this (and other evidence) is most likely to
arise from peer e¤ects on the information that individuals have and not their
preferences.
We are very grateful to teachers from the Universidade de São Paulo for their support and help
in the application of the survey. We thank Renata Rizzi for helping with the data collection, and Lori
Beaman, Annemie Maertens, Je¤ Prince, Francesco Trebbi and participants from the RES Women
Committee Meeting for comments.
y
Insper Institute of Education and Research. Assistant Professor. Department of Economics. Email: [email protected]
z
King’s College London.
Professor.
Department of Political Economy.
E-mail:
[email protected]
x
University of East Anglia. Assistant Professor. Department of Economics. E-mail: [email protected]
1
1
Introduction
People often behave alike when they know each other well. Friends, for example,
frequently vote for the same party, send their children to similar schools, choose the
same types of vacations or enjoy eating at certain restaurants and not others. Groups
are formed by such commonalities and they pose a fundamental question for social
science. Do such commonalities arise because people with prior preferences for ‘x’
naturally associate with fellow ‘x’seekers and share information or does membership
of the group encourage conformity because the psychological dynamics within a group
are such that individual preferences become more alike? This is the question that we
address in this paper with a natural experiment, focusing on political behavior.
The question matters because much in economics and liberal political theory turns
on taking individual preferences as given. The appeal, for instance, of the Pareto criterion in welfare economics and of the ‘will of the people’in politics depends on being able
to identify individuals with their preferences and this becomes problematic if an individual’s preferences change with those of their peers. The question, however, is di¢ cult
to answer. To control adequately for possible prior commonalities, common shocks and
the role of information transmission within a group, and so identify whether there is
a distinct peer e¤ect on individual preferences, is not easy. This is why experiments,
where the scope for such control is often greater, are attractive. The laboratory experimental evidence, however, is mixed on this general question. For example, Hung and
Plott (2001) interpret the evidence from their information cascade experiment as telling
in favour of information transmission and against preference change in the explanation
of behavioural conformity. But, the evidence on the unpredictability of music bandwagons in the Salganik, Dodds and Watts (2006) experiment is di¢ cult to reconcile
with information transmission alone. In this paper, we report on a natural experiment
where we attempt disentangle the contribution of prior commonalities and the possible
information transmission e¤ect within a group from the possible in‡uence that peers
have on other individuals’speci…c political preferences.
In particular, we consider whether there is evidence of peer e¤ects on two types of
individual political behaviours. One is an individual’s substantive political identi…cation on a left-right spectrum and the other is on an individual’s engagement with the
process of politics that is revealed by their acquisition of information on candidates
in an election and their willingness to vote in an election.1 Where there is evidence
1
Given the Public Choice insights with respect to ‘rational ignorance’and the ‘paradox of voting’, an
individual willingness to acquire information and/or vote is often regarded as indicating that individual
2
that peer’s political identi…cation and/or engagement a¤ects individual political identi…cation and/or engagement, we exploit aspects of the experimental design to consider
whether it arises from a peer in‡uence on the information that individuals have or over
their preferences.
There is a large literature on peer e¤ects in politics.2 The speci…c evidence on peer
e¤ects on political identi…cation is also mixed. Some studies …nd evidence consistent
with the claim that people follow their peer’s political a¢ liations (Sinclair, 2009; Beck,
2002; Kenny, 1994), others …nd no association (MacKuen and Brown, 1987). But much
of this is based on correlations that are subject to selection biases: that is, the correlations could arise from people with shared prior commonalities naturally being drawn
together. We address this di¢ culty in the experiment by exploiting the fact that our
data consists of freshman students who have been randomly divided between di¤erent
class groups for the introductory courses in their chosen major subject. This means
that the characteristics of the peers in a person’s class group should be independent of
his or her own characteristics. We interview students twice in an election year (before
the presidential campaign and after the Election). To test for peer e¤ects, we examine
how and whether their identi…cation and engagement in the second survey correlates
with their classmates’initial political orientations and engagement.
There are other studies that use an experimental or quasi-experimental framework
for the same reason. For example, Sacerdote (2001), Carrell, Hoekstra and West (2011)
and Lyle (2009) use data on randomly assigned networks to identify peer e¤ects on, respectively, student performance, physical …tness and workers’productivity.3 The closest
has some kind of a ‘social preference’that is revealed by this kind of engagement with politics. Thus
we examine political behaviours where there are both personal and social preferences that are plausibly
in play.
2
Many studies investigate how individuals’behavior is associated with the behaviour or characteristics of their household members (Nickerson, 2008), people who live in the same geographical and
residential area (Cho, 2003; Cho et al.; 2006, Huckfeldt and Mendez, 2008, Huckfeldt et al.,1995,
Huckfeldt and Sprague, 1987), housemates (Klofstad 2009; 2010), discussion partners (Huckfeldt,
2007, Mutz, 2002a, 2002b; Gerber et al., 2012), co-workers (Mutz and Mondak, 2006) or facebook
friends (Bond et al 2012).
3
In political studies, this strategy has also been used by Klofstad (2009; 2010). By examining
di¤erences among randomly assigned college dormitory roommates, he …nds a positive relationship
between the level of discussion of politics among freshmen and participation in civic events. Our study
di¤ers in important respects. As will be discussed in Section 2.3, we use student characteristics and
preferences rather than behaviour in our peer variables. This avoids the criticism that something
like the level of discussion about politics among roommates is not random, but determined by the
combination of individuals’ unobservable characteristics such as personality or preferences that are
correlated with civic participation. In this sense, we have a stronger case for causality. Secondly, we
look at di¤erent peer characteristics to understand the relevant mechanisms of peer in‡uence. Lastly,
our study investigates e¤ect of groups, namely classmates, rather than the e¤ect of one roommate.
3
to our study are those natural and …eld experiments that have been used to quantify
peer e¤ects on voting turnout (Funk, 2010; Panagopoulos, 2010; Gerber, Green and
Larimer, 2008, Nickerson, 2008.) Their …ndings are consistent with the fact that voting
is contagious in social circles as people respond to social pressure by becoming more
likely to vote. But little is known about the mechanism producing conformity in this instance. Does it arise because individuals become better informed about political choices
through interaction with peers and so become more inclined to vote? Or do peer preferences for political engagement strengthen what would otherwise be weak individual
preferences for political engagement?4 The di¤erence matters for the reason sketched
above: the ‘will of the people’becomes a less appealing justi…cation for democracy when
individual ‘will’‡oats with those of others. Our experiment enables us to distinguish
between these possible explanations of peer e¤ects on likelihood of voting, as well as
other aspects of political behaviour.
We …nd no evidence that peer political identi…cation in‡uences either individual
political identi…cation or political engagement. In this respect, we …nd no reason to
doubt the presumption that individual preferences can be taken as given. Interestingly,
when we relax the controls for prior commonalities among the members of a group, we
…nd an apparent peer political identi…cation e¤ect on individual political identi…cation.
This suggests that the failure to control fully for prior commonalities can, in practice, be
a serious problem: it can lead to misleading inferences over the sensitivity of individual
behaviour to peers.
We do …nd evidence, however, of a peer engagement e¤ect on individual political
identi…cation, willingness to vote and the political knowledge among those individuals
who are initially least engaged. In particular, those individuals who are initially least
engaged become more likely to vote and their political identi…cation is more likely to
move to the centre (away from the extremes) as the engagement of their peers increases.
This might seem troubling for those who take preferences as given. But, it is also
consistent with information transmission within the group because there is evidence
that the least engaged become better informed, particularly about the candidates they
are initially not inclined to vote for, and this is despite the fact that they are not more
inclined to acquire information. The fact that the more informed are typically not
4
These are open questions. Claudine Gay (2009) discusses the lack of knowledge about the subject
in putting forward her perspective about the Future of Political Science: "We know relatively little
about how contexts in which individuals are situated shape politically relevant beliefs and opinions,
and subsequently, behavior: What features of context matter? What are the mechanisms of contextual
in‡uence? What is the range of behaviors and attitudes a¤ected? A full and compelling account of
the political life of the mass public is impossible without greater attention to these questions."
4
a¤ected by their peers in this way also suggests an information rather than a preference
link between individuals in the group underlies peer e¤ects.
In the next section, we describe the natural experiment on freshman students at
Brazil’s largest university, explain the data and we set out the model that we use for
identifying peer e¤ects in Section 3. Section 4 presents the estimates of peer e¤ects.
Section 5 discusses these results and we conclude in Section 6.
2
Experimental Design and Identi…cation of Peer
E¤ects
2.1
Overview
Our subjects are freshman students at the Universidade de São Paulo (USP). The
move from high school to university marks a natural transition to adulthood where
new networks are formed. USP is also the largest university in Brazil and the freshman
students are randomly allocated to classrooms, and hence classmates. As result, these
classes plausibly represent the creation of new peer groups for the incoming students.
Our strategy was to sample these freshman students early in the academic year and
prior to the commencement of a Presidential campaign to establish prior values of the
individual variables relating to preferences for political a¢ liation and engagement. For
each individual, we calculate peer e¤ect variables for two key measures relating to the
political engagement and political a¢ liation. We, then, re-survey the sample at the conclusion of the Presidential election and test whether the individual political a¢ liation,
engagement and knowledge at this later date correlates with the peer variables.
The choice of freshman students who are entering during a Presidential election
year as subjects is a key aspect of the experimental design. The fact of the election
makes the transition to adulthood particularly salient because voting is compulsory for
everyone aged 18 or above in Brazil and the campaign that occurs between the …rst
and …nal sample of individual variables is a natural political event that might cause our
subjects to think about politics and so become exposed to peer e¤ects, if there are any.
There are also strong grounds for supposing that the social life in classrooms is an
appropriate environment to measure peer e¤ects. USP freshmen have all their introductory lectures with the same group of classmates during their …rst term in university
(when we …rst interview them). They have at least two lectures together per day5 and
5
Students in morning courses have two lectures per day, from 7:30 to 11AM, while students in
5
they interact outside the class with each other though academic activities such as study
groups and joint course projects. In addition, there are fewer alternative university
peer groups than is typically the case at UK and US universities because most students
are local and live at home (74%). Classmates are the …rst group of students they meet
in college and they are a relatively large pool of possible friends (the average size of a
class is 54 students). In short, between our surveys, students became friends, interacted
in classes, and were exposed to a presidential campaign that made politics salient for
discussions within social circles.
2.2
The Subjects and Method of Data Collection
USP has approximately 86,187 students enrolled and o¤ers 229 undergraduate and
graduate courses. To be enrolled, undergraduate students have to complete secondary
education and pass an entrance exam ( “Vestibular”), which is USP-major speci…c and
runs once a year. USP is a public university, that is tuition-free, and it is one of the
most prestigious universities in Latin America. For these reasons, the USP entrance
exam is highly competitive: for instance, in 2011, the number of applicants was 138,888
and the year’s enrolment was only 10,202.6
Our data come from 2010 freshman students enrolled in speci…c subject majors:
architecture, business administration, economics, history, law, literature, mathematics,
physics, and sociology. For these majors, USP admits more than 180 students per
year and divides the freshmen into at least two classes for the introductory courses.
While students obviously choose their subject major, they cannot choose their class
assignment: it is based either on alphabet order or a university algorithm.7 Since the
initial process of allocating students to classes is largely random, our classes and the peer
variables should be free from the more obvious sources of selection bias. Nevertheless,
we check for random assignment in section 4.1.
The same survey procedure was used in all classes. An interviewer entered the
evening have lectures from 7:30 to 11PM.
6
Only those students with top scores on the admission entrance exam are accepted. The level
of competition varies by major of choice. For example, in the 2011 USP admission exam, 13,545
individuals applied to study Medicine and were competing for one of the 120 vacancies available. On
the other hand, 260 individuals applied to study Mathematics, competing for one of the 112 places
available (Fuvest, 2011). More information about USP follows here: http://www4.usp.br/.
7
The speci…c number of classrooms depended on students’ major class, which is a combination
of major and the time of the day that students take classes – namely, 2 (for students enrolled in
economics-day, economics-night, history-day, mathematics-night, physics-day, physics-night, business
administration-day, law-day and sociology-night), 3 (for history-night), 4 (for architecture-day, , lawnight, sociology-day), 7 (literature-day), and 8 (literature-night).
6
classroom about 15 minutes before the end of a class, read an introductory script
aloud, and distributed the questionnaires to all students. Lecturers also contributed
by asking that attention and consideration be given to the survey. Students, then,
had 10-12 minutes to …ll out, individually the questionnaires. The survey was titled
“Young Adults’Political Behaviour”and the contact details of the authors were given
for further information. The instructions made clear that students should answer the
survey individually. In every class, four types of questionnaires – containing exactly
the same questions but in di¤erent order –were randomly distributed to students (to
encourage individual answering). The large majority of students agreed to answer
the survey (in a few classrooms, one or two students failed to return the …lled-out
survey), and 95.54% of the respondents declared that they had answered questions in a
serious manner. The questions related to individual demographics, political knowledge,
political identi…cation, media consumption and their parents’political commitments.
The …rst wave, pre-election, was administrated during April 2010 (henceforth, referred to as t 1). The questionnaires were collected before the formal entry of all
candidates in the race or of their running mates (in June); the beginning of the TV
presidential campaign (in July) or any of the three debates on TV (in August and
September). To assess the level of discussion regarding the 2010 Presidential Election
in society as a whole at these times, we conducted an engine search in the one of the
largest newspapers of São Paulo, O Estado de S. Paulo, seeking stories containing the
word “elections”.8
F igure1
During the weeks in which this wave of the survey was conducted, the word “elections”appeared in 65 stories. This number substantially increased in July, an increase
which coincides with the start of the compulsory television electoral advertising,9 and
they were 134% higher in September (month preceding the election, with a peak of
150 stories). The newspaper articles in April were mostly about possible and likely
candidates entering the presidential race rather than their policy issues or candidates’
government agenda. Consistent with this, opinion polls registered with the Supreme
Electoral Court show that the volatility of intended votes for the two main candidates
8
However, we only considered stories referring to the 2010 Brazilian presidential elections, checking
one by one.
9
In Brazil, television channels are required to broadcast candidates’advertising during weekdays,
including at least a half-hour on prime time, starting around three months before election day (Da
Silveira and De Mello, 2011).
7
– Dilma Rousse¤ and José Serra – increased abruptly by July (see Figure A1 in the
Appendix). This suggests that people tended to form opinions and discuss politics more
enthusiastically closer to election: that is, from July on. Based on this evidence, we take
the …rst wave of the survey as supplying information on pre-determined characteristics.
The second wave of the survey was administrated just after the …rst round of presidential elections, during October 2010 (henceforth, t). Students were asked the same
questions as in the …rst wave. The data in the …rst survey consisted of 1,593 student
responses from 48 classes, the data in the second wave had 1,103 student responses
from 39 classes. Our panel sample consisted of the students that had responded to
both surveys, a total of 635 students.1011 This is the main sample used in the analysis.
It represents 39.8% of the initial sample. Two things should be noted about this. First,
the peer variables for these individuals are calculated on the basis of the larger initial
survey of relevant individuals. Second, the panel sample has many similarities with
USP students.12 We test for whether the attrition is in any sense unbalanced or not
random so as to bias results. We do this in two ways.
First, we investigate if there is any correlation between abstention in the second
survey and our peer variables. We investigate this association across students within a
major-class (e.g. comparing the behaviour of students enrolled in Economics-evening,
but that are assigned to di¤erent classrooms).13 This is the level of randomization,
10
The panel was identi…ed based on responses about names, date of birth, and enrolled major. For
a few cases, we also conducted checks on students’handwriting across surveys.
11
The change in numbers between the two surveys partly occurred because the second wave of the
survey was conducted in fewer classrooms. Although all contacted teachers agreed to allow us to
survey their students during the …rst wave, Law and Architecture lecturers were conducting reviews or
midterm exams during the second wave of survey. For this reason, many refused to let us conduct the
survey. Another reason for the lower number of observations in the panel is that some students did not
provide their names in the second wave and hence, we could not link their answers to the ones in the
…rst survey – this occurred in 17.3% of cases (=192/1103). Finally some students missed the lecture
on the day the survey was administered and among those that claimed to have answered a political
survey before, 89.1% were found to be in the panel.
12
We compared the characteristics of our sample with publicly available administrative records for
freshmen classes. The results are presented in Table A1 in the Appendix. In general, students in our
sample are less likely to come from lower socio-economic background than the universe of freshmen
students, re‡ecting that more a- uent students are more likely to attend classes (recall that USP is
tuition-free). This is the population more exposed to classmates and to peer e¤ects. It is important
to note that such socio-economic selection of students is observed both in the panel and among all
students observed in the …rst survey.
13
We estimate regressions of the following for:
Prob(Be on the panel ) =
The coe¢ cient
Peer Variable + major
xed e ect
is not statistically signi…cant for any peer variable, with p-value of at least 30%.
8
in which we conduct our main analysis. We …nd no association. According to our
results, for example and importantly, variations in the proportions of classmates that
self-declare right-wing or those have a partisan parent in t 1, are unlikely to cause
abstention in the survey in t.
Second, we simulate random panel groups. For each classroom, we randomly drew
a sample without replacement, with the same size as the one observed in the panel.
We repeated this process 1,000 times to obtain an empirical 90% con…dence interval for
each major-class. We then compared the characteristics in the actual panel with the
simulated ones. Figures A2 in the Appendix shows the average values of students’characteristics for those observed in the actual panel as compared to the con…dence interval
that was based on the simulated panels. For most of the demographic characteristics,
the panel data falls inside the con…dence interval. Only two student political identi…cation average values –centre-oriented and left-wing-oriented –fall outside the con…dence
interval, and for only two cases (sociology-day and law-day which account for only 6%
of the observations); and only one engagement variable, the probability of casting an
invalid vote, falls outside the con…dence interval for two major-classes. The attrition,
therefore, does not alter the sample much in these respects, but, to be sure that there
is no biasing e¤ect, we additionally control for students’ observable characteristics in
the main regressions.
2.3
Variables
Table 1 gives descriptive statistics on the individual variables at t 1 and t and
the peer variables. The pre-determined individual characteristics (observed in the …rst
survey) are set out in Panel A. They relate to the usual demographics, whether individuals have a partisan parent and whether they intend to vote. Although voting is
compulsory, there is an option of voting for no one: that is, in e¤ect, abstaining from
voting by casting an invalid vote. We use this as our measure of willingness to vote.14
In addition, to their declared political identi…cation (left-wing, center and right-wing),
we asked students to cite the three most relevant socioeconomic problems among …fteen alternatives (e.g. unemployment, high interest rates, taxes, public health system,
poverty in the city, tra¢ c jams, violence and corruption in the government). Answers
to this are possibly less likely to be misreported than the self-reported political identi…cation and so we use these responses to construct a revealed political identi…cation
index, as set out in (1) below.
14
This is typically interpreted (e.g. Maringoni, 2006) as a form of protest voting.
9
RightW ingIndext
1i
=
"
3
X
Choiceipt
p=1
1
#
RightW ingp =3
(1)
where Choiceipt 1 is an indicator of individual’s i choice of problem p in survey t 1
and RightW ingp is the proportion of right-wing-oriented individuals (other than in i)
who chose problem p in t 1.15
The summary statistics for the peer variables are given in Panel B. The peer political
identi…cation variable for each individual is based on classmates’direct responses to the
political identi…cation question (whether they identify as right-wing oriented). The peer
political engagement variable is the proportion of classmates that declared that their
parent prefer a particular party. We call this the partisan parent peer variable. Those
with partisan parents reveal in the …rst survey a greater willingness to vote and watch
the campaign on television than those who do not (see Table A2 in the Appendix). This
is perhaps not surprising since politically committed parents seem likely to encourage
political engagement among their children through discussion and the like. We use this
peer variable to capture political engagement in case individuals have any tendency to
misreport their level of political engagement (since it seems likely that if there is such
tendency it will be much weaker when reporting on one’s parent’s than on one’s self).16
It is important to note that although one might expect that students in a same
major-class are largely homogenous, there is sizable variation in both peer variables
within major-class (see Table A3 in the electronic Appendix) and this is an important
ingredient for the identi…cation of peer e¤ects.
The outcomes (observed in the second survey) are summarized in Panel C. We
use information on individual consumption of the media at this time and individual
political knowledge. The latter comes from a quiz containing the same number of
15
These indexes can vary between zero (i.e., no right-wing student made any of the same choices of
i) and one (i.e., all right-wing students made the same choices of i). Lef tW ingIndext 1i is analogous.
To illustrate, suppose that all right-wing individuals (as self-reported in t-1) but i were concerned
only about violence, corruption and taxes. If the individual i chose violence, public transportation
and environment as his main concerns in t, then the value of his RightW ingIndext 1i is 1/3 (=[1
x1]+[0 x 1]+[0 x 1]/3). A higher value of the right-(left-)wing index indicates more similarity with
the concerns of right-(left-)wing students. Individuals’ ideologies predict their ideology indexes. We
conducted regressions to determine that indicators for right- and left-wing orientation are predictors of
the ideological indexes. Right-wing-oriented individuals have a higher right-wing index (28% standard
deviation) and lower left-wing index (29% standard deviation), whereas left-wing individuals have a
lower right-wing index (19% of standard deviation) than others.
16
When we formally instrument the pre-determined interest in politics using this partisan parent
variable, we obtain, e¤ectively, the same results. These results are not shown in the paper, but are
available upon request.
10
analogous questions about each of the main Presidential candidates, Dilma Rousse¤,
Jose Serra and Marina Silva.17 We construct two knowledge variables that take account
of the voting intentions at the time of the …rst survey. The variable, ‘Mistakes on
Own Intended Candidate’, computes the proportion of mistakes in t made about the
presidential candidate the student intended to vote for in t 1. Similarly, we create
the variable ‘Mistakes on Remaining Candidates’ which computes the proportion of
mistakes made about the other presidential candidates. As a more general measure of
(dissimilarity of) knowledge about the candidates, we consider the sum of the pairwise
di¤erences in mistakes made about candidates Assymetric Mistakes’.18 A higher value of
this variable re‡ects more asymmetric knowledge and less knowledge about candidates.
Lastly, we repeated questions on political identi…cation and socio-economic problems
in t, and construct measures of preferences, like in t 1.19
T able1
3
Identi…cation of Peer E¤ects
Following the literature on the identi…cation peer e¤ects through experimental
techniques (Lyle 2007; Sacerdote 2001), we assume students’outcomes are a function
of individual and peer characteristics, as in (2).
t
Ymci
=
+
t 1
1 Xmci
t 1
2X c i
+
+
t
3Y c i
+
t
mci
(2)
t
, is the outcome at time t of individual i, enrolled in major-class
The variable Ymci
17
More speci…cally, there were three open-ended questions about their previous political position,
their party experience and their running mates, as well as four multiple-choice questions about policies
previously implemented or supported by the candidates.
18
This is de…ned as:
AsymmetricM istakes = jMRousef f
MSerra j + jMRoussef
MSilva j + jMSerra
MSilva j
where MC stands for the number of mistakes made about each one of the three main candidates.
19
The political index in t is de…ned as:
RightW ingIndexti =
"
3
X
#
Choiceipt RightW ingp =3
p=1
where Choiceipt is an indicator of individual’s i choice of problem p in survey t and RightW ingp is
the proportion of right-wing-oriented individuals (other than in i) who chose problem p in t 1.
11
t 1
m, allocated to classroom c; Xmci
corresponds to own individual’s pre-determined chart
acteristics. The variable Y c i represents the average behavior of students in classroom
t 1
c (excluding i) by t and X c i are average characteristics of students in classroom c
(excluding i), at time t 1.
Using Manski’s (1993) outline, 3 and 2 correspond respectively to endogenous –
that represents contemporaneous and simultaneous in‡uence of peers –and exogenous –
a sole in‡uence of classmates on individuals –peer e¤ects. As explained by Lyle (2007),
the error term tmci can be decomposed into three terms ( tmci = t1ci1 + t2c + t3mci ), where
t 1
t
t
1ci represents an unobserved selection term, 2c represents common shocks and 3mci
represents, a standard error term. In a non-random assignment setting, we could expect
t
t 1
t 1
a correlation between 1ci
and the peer variables (Y c i and X c i ), as students’choice
of whom to socialize with are based to some extent on individuals’ tastes, which are
unobservable to the researcher. This could lead to a possible bias in the estimates for
2 and 3 .
A related issue is that members of the same social group could be exposed to common
external shocks/in‡uences over the year (e.g., reading the same newspapers and participating in the same political events), thus leading to a positive bias for the estimates
of 2 and 3 . This possibility is less likely when the initial allocation of individuals
to classes is random (and we will show in Section 4.1, that there is no evidence of intentional selection). Further, since all students in the same major-class take the same
classes and are exposed to the same college environment, it seems plausible to assume
they are exposed to similar sets of external in‡uences over the election year. One important quali…cation, however, is that some shocks might be particular to students in some
classroom: for instance, the exposure to an instructor with extreme political views. An
underlying assumption is that the in‡uence of punctual shocks vanishes on aggregate
when considering all external shocks at a level as …ne as the classroom. This hypothesis
is particularly important when estimating contemporaneous peer e¤ects ( 3 ), as comt
mon shocks might lead to some correlation between t2c and Y c i . For example, Lyle
(2007) demonstrates that common shocks represent a confounder for the estimate of
contemporaneous peer e¤ects ( 3 ) even in the presence of a setting with random assignment, for the reason discussed above. Di¤erently, common shocks over the election year
are unlikely to be correlated with the distribution of students’pre-determined charact 1
t 1
teristics across classrooms (X c i ) at the major-class level;20 therefore, E[X c i ; t2c ] = 0.
t
t
For this reason, we take the average of Ymci
across classmates and obtain (3), with Y c i
20
i.e., since these are determined by the realization of the past classroom assignment lottery.
12
t
. Substituting (3) in (2) and rearranging, we obtain (4), which is
as a function of Ymci
the reduced form to be estimated and depends only on predetermined characteristics.
t
Yc
i
= +
t
=
ymci
t 1
1 Xmci
0
+
+
t 1
2X c i
t 1
1 Xmci
+
+
t
3 Ymci
t 1
2X c i
+
+ ! tmci
t
mci
(3)
(4)
As a result, under random assignment, the coe¢ cient ( 2 = 12 + 3 2 ) captures peers’
3 3
in‡uence since it is free from a correlation with the error term. This peer e¤ect is a
function of both endogenous and exogenous peer structural parameters and these e¤ects
are not disentangled in this paper.
4
4.1
Results
Random Assignment
The key identifying assumption of our study is that, conditional upon the majorclass of study, students were randomly assigned to classrooms. USP uses a randomizing
procedure and we now check whether it had this e¤ect. We perform the test proposed
by Sacerdote (2001), regressing the peer characteristics of interest on the corresponding
average of their peers. Since classrooms are small, even under random assignment, a
negative correlation might be expected.21 To control for this, we applied the correction
proposed by Guryan, Kroft, and Notowidigdo (2009). They conclude that it su¢ ces to
include in the typical test the average value of the characteristic being inspected among
all students in the same major-class, excluding individual i (z tm 1i ). The modi…ed test
corresponds to:
t
zmci
=
+ z tc
1
i
+ z tm 1i +
t 1
m
+ "tmci
(5)
t
where zmci
is the outcome of individual i, enrolled in major-class m, who takes
classes in the …rst college term in classroom c. The variable z tc 1i is a peer measure
based on the average characteristics of classmates while attending college in the …rst
term, excluding himself.
21
As explained by Guryan, Kroft and Notowidigdo (2009), this stems from the fact that individuals
cannot be their own peers. ". . . [T]he sampling of peers [in classrooms] is done without replacement –
the individual himself is removed from the ‘urn’from which his peers are chosen."
13
If peers are assigned randomly, the coe¢ cient should be zero. The results for the
modi…ed Sacerdote test are presented in column 1, row 1 and in column 2, row 2 of
Table 2 in bold. The coe¢ cients for the peer variables are not statistically di¤erent
from zero, thus supporting the assumption that selection is not underlying our main
results.22 We also checked whether changes in the proportion of peers with partisan
parents and in the proportion of right-wing peers were systematically correlated with
other predetermined characteristics, estimating (5), but using as dependent variables
demographic characteristics and political behavior variables. The results are reported
in Table 2, not in bold (rows 3-12, row 1 column 2; and column 1 row 2) .
T able2
The peer variables are not related to students’ media consumption of politics or
intention to invalidate their votes, which suggests that no di¤erence exists in these predetermined political involvement across students assigned to di¤erent classes. There is
a negative association between the proportion of classmates with a partisan parent and
students’propensity to identify themselves as right-wing. This association might arise
for two reasons: (i) luck or (ii) because some students felt afraid of declaring themselves
to be right-wing-oriented. The latter would be worrying but the same correlation, which
would be expected under (ii), is not observed among students’ propensity to declare
themselves to be right-wing-oriented and the proportion of classmates that are rightwing. Nevertheless, we also include classmates’ average characteristics as additional
controls in the main regressions.
4.2
Peer Political A¢ liation E¤ects on Individual Political Af…liation
Table 3 provides the estimates of a version of equation (4)23 using as dependent
variable, individual political identi…cation and focusing on the in‡uence of peers right
wing identi…cation. The dependent variable is in all cases except column 7, an indicator
22
It is important to note that these results hold on average for students assigned by di¤erent rules –
namely, random algorithm or alphabetical order. There is evidence in the US that …rst names convey
individuals’demographic characteristics (Bertrand and Mullainathan, 2004). To understand whether
a similar pattern was a¤ecting our exercise, we looked for di¤erences in classmates’ characteristics
according to the …rst letter of their …rst name (i.e., classroom assignment). In results not shown in the
paper, we …nd that classmates’characteristics do not di¤er by name allocation for any of the observed
characteristics.
23
We add major class …xed e¤ects (
t 1
m )
to this equation.
14
for whether the individual self declares as right-wing in t; and in column 7, we use the
implied index of right-wing sympathy that emerges from the individual’s ranking of
issues (de…ned in (1)). Each column reports a separate regression that di¤ers according
to the controls.
T able3
Our preferred speci…cation is in column 4 and 7 where we control for possible sources
of selection bias (including in the regressions indicators for gender, race, age, income,
mother education, political identi…cation at t 1, major-class …xed e¤ects and the
proportion of classmates that declared to have a partisan parent in t 1). We do
not detect any peer e¤ect. The coe¢ cient on the proportion of right-wing classmates is
practically equal to zero and it is not statistically signi…cant. The results for regressions
using individual left-wing identi…cation as dependent variable mirror the ones reported
here and, for reasons of space, are not presented.
We considered possible more complex peer political identi…cation e¤ects. First, we
test for an interaction term between the peer right-wing identi…cation and an indicator
for whether the student self-declared right wing in t 1 (column 5). The purpose was to
check whether peer e¤ects worked speci…cally by reinforcing students’pre-determined
preferences. Second, in column 6, we add to the regression a peer variable that is
the proportion of classmates that both self-declare right wing and have a partisan
parent in t 1. This is to test for the possibility that individuals are a¤ected by the
preference of their engaged peers. Again, we …nd no statistically signi…cant e¤ects. We
do …nd, however, that individuals’own pre-determined party political identi…cation are
important in explaining political identi…cation in t (Table 3, not in bold).
Result 1: There is no evidence that the political identi…cation of an individual is
a¤ected by the political identi…cation of his or her peers.
In comparing these …ndings with those in Table 3, columns 1 and 2, we check
whether this result is sensitive to the control for selection biases. We …nd it is. If there
is no control for individual demographic characteristics and/or if there is no control
for choice of subject major, the peer political identi…cation variable becomes signi…cant
and positive. In other words, in the absence of control, it appears that peers do a¤ect
an individual’s political identi…cation.
Result 2: The failure to control for selection biases creates the (false) impression
that an individual’s political identi…cation is in‡uenced by the political identi…cation of
15
his or her peers.
4.3
Peer engagement e¤ects on individual engagement and political a¢ liation
Table 4 presents the results of the regressions that test for a peer partisan parent
e¤ect on individual engagement, political a¢ liation, knowledge and consumption of the
media. Each entry reports a separate regression result for the peer coe¢ cient, using
as dependent variable, measures presented in the rows. All regressions include controls
for gender, race, age, income and mother education indicators, political identi…cation
at t 1, a dummy for whether the student declared to have a partisan parent in t 1,
major-class …xed e¤ects (to take into account the classroom random assignment) and
the proportion of classmates that declared to be right-wing oriented in t 1.
T able4
In column 1, we report results for the whole sample. There is evidence of a peer
engagement e¤ect on willingness to vote, on political identi…cation (where it appears to
encourage movement to the centre, notably from the left) and on political knowledge
of candidates other than the one’s preferred one, but not on consumption of the media.
Next, we supply additional insight into this partisan parent peer e¤ect by looking at
di¤erential peer e¤ects based on pre-determined individual engagement (self reported
parent a¢ liation and media consumption in t-1). In columns 2 and 3, we repeat the
analysis for distinct sub-samples of students: those that in t 1, declare to have a
partisan parent and those who do not have any partisan parent. The partisan parent
peer e¤ects are only statistically signi…cant in regressions reported in column 3. Recall
that individuals with partisan parents tend to consume more information initially and
are less likely to cast invalid votes. They are the ones who are more engaged with
politics in t 1 and it seems that they are not in‡uenced by their peers in these
respects (column 2); whereas those who were less engaged with politics at t 1 are
in‡uenced by their peers in these respects (column 3).
In Columns 4 and 5 split the sample into those who are low media consumers in t 1
and those who are high media consumers.24 Consistent with results in columns 3 and
24
The variable consumption of media indicates the number of days the individual follows the news
on TV, newspapers and Internet. We summed the number of days across all media (TV, newspapers
and Internet) and de…ne High media consumers as those that consume above the median (10 days of
16
4, the partisan parent peer e¤ects appear to be strongest for the low media consumers
(e.g. e¤ects on political identi…cation and willingness to vote are signi…cant at 95%
level only for this sub-group) but in general they are less apparent (e.g. there is no
signi…cant peer e¤ect for either sub-group’s political knowledge).
Result 3: Individuals who were initially less engaged with politics are the most
a¤ected by the level of engagement of their peers. This peer e¤ect encourages voting,
a movement of political a¢ liation to the centre (away from the extremes) and political
knowledge (speci…cally of candidates that they were not planning to vote for at the
outset).
Result 4: There is no evidence among the least engaged that an individual’s consumption of the media is a¤ected by the level of political engagement of his or her
peers.25
5
Discussion
Our results are important in two respects.
First, we contribute to the debate in the literature over whether peer political identi…cation a¤ects individual political identi…cation. There is mixed evidence on this: e.g.
Mackuen and Brown (1987), on one side and Kenny (1994), Beck (2002), and Sinclair
(2009), on the other suggesting that there is an in‡uence. We …nd no evidence of such
an e¤ect (Result 1). Crucially, the earlier studies rely on correlations that do not control systematically for prior commonalities. In contrast, we are able through the use of
the experimental method to control for these selection biases, by comparing behaviours
among individuals randomly assigned to di¤erent peer groups. This is important not
only because, once we control for these sources of similarity, we …nd no peer political
a¢ liation e¤ect, but also because our study suggests that the failure to control fully
for these selection biases is, in practice, material (Result 2). Looking only at simple
correlations, we …nd that the political a¢ liation of a person’s peer group is a good
predictor of that person’s political a¢ liation. For example, an increase of 10% in the
right-wing class mates would appear to increase by 9.1% the chance of an individual
media or more) and as Low Media consumers those that consume less than the median (less than 10
days).
25
There is some evidence, however, that the initially more engaged reduce their consumption of the
media (column 2, row 8 and column 5, row 9)
17
declaring a right wing political a¢ liation. However, once we control for selection biases,the predictive power of the peer political identi…cation disappears, yielding a very
di¤erent conclusion.
Second, we …nd evidence of a di¤erent kind of peer e¤ect that is largely new to
the literature: individuals who are initially less politically engaged are a¤ected by the
level of political engagement among their peers. In one respect this echoes a result in
the literature with respect to social contagion in voting (Gerber, Green and Larimer
2008, Nickerson, 2008), but our result is more complicated in two respects. One is that
the peer e¤ect we identify is more speci…c in the sense that those who are initially less
engaged are a¤ected by their peers but others are generally not. The other is that
the peer e¤ect is more pervasive in the sense that not only is voting decision among
the initially less engaged in‡uenced by peers for this sub-group, so is their political
identi…cation and political knowledge (Result 3).
One possible interpretation of these …ndings is that the preferences among the initially less engaged change through interaction with the more engaged. The group
dynamics are such that people within the group become more alike in their preferences.
This interpretation, however, is di¢ cult to reconcile with other aspects of our …ndings.
For instance, if individual preferences bend with those of their peers, why is there no
evidence of this with respect to the in‡uence of peer political identi…cation (Result 1)?
Surely, one might expect such preference osmosis to occur more readily between peer
and individual political identi…cation than between peer engagement and individual
political identi…cation. Further, if there is preference osmosis at work, why does it only
operate for those who are initially less engaged? Finally, while a change in individual
preference for political knowledge might explain why some individuals become better
informed, this is di¢ cult to reconcile with the absence of any such peer e¤ect on media
consumption (Result 4).
The alternative interpretation that avoids these di¢ culties is that the least engaged
learn something about the candidates in the election from those who are more engaged
through social interaction with their peers. This explains why the e¤ect only works
one way since interacting with someone who knows less does not corrode the knowledge of those who know more. It also explains why the less engaged become better
informed without actually becoming more inclined to engage explicitly in information
acquisition through more media consumption (i.e. the improvement is a by-product of
social interaction). Being better informed can also explain why they might be more
inclined to vote since they are now in a better position to decide between candidates
(see Lassen, 2005, and Gentzkow, 2006, for evidence that information causes political
18
participation). Finally, the fact that they speci…cally learn about candidates who they
were not initially inclined to vote for makes it more likely that their a¢ liation will move
to the centre of the political spectrum (because, in e¤ect, of regression to the mean).
Again this is consistent with other …ndings in the literature (e.g. see Banerjee et al.,
2010).
In short, our results are important because they suggest that there are signi…cant
peer in‡uences on individuals but they are not the ones that encourage worries for
democracy on grounds of ‘group-think’. Indeed, the reverse is the case. An individual’s
political identi…cation is not a¤ected by that of their peers in our experiment. Peer
political engagement does a¤ect individual engagement and political identi…cation, but
only among those who are initially least engaged. While these peer e¤ects could arise
because individual preferences exhibit ‘group-think’malleability, they, more plausibly,
re‡ect information transmission between individuals about the political process. In so
far as this is right, then these peer e¤ects, far from damaging democracies, are likely to
be bene…cial because democracies seem likely to function better when people are both
better informed about candidates and more inclined to vote on the basis of this better
knowledge.26
6
Conclusion
Using a natural experiment on young adults, we examine whether peers in‡uence
individual preferences over political a¢ liation and political engagement. The particular
question is important because the appeal of democracy in terms of its responsiveness to
the ‘will of the people’depends on being able to identify individuals with their preferences. If an individual’s preferences depend on his or her peers, then individual ‘will’in
this sense has slipped its anchor in the individual. But it is also an instance of a general
question that has wider importance for social science. For instance, much of economics
takes individual preferences as given, the starting point for analysis, so to speak, and
it would be equally damaging here to discover that an individual’s preferences ‡oated
26
There is a literature concerned with how individuals respond to heterogeneity of substantive belief
in their group that we are also able to address in this experiment. Mutz (2002a), for example, suggests
that people become confused in the presence of disagreements and so tend to participate less politically
when in a heterogeneous group. In a di¤erent vein, Mutz and Mondak, (2006) and Mutz (2002b)
…nd that heterogeneous groups encourage tolerance. We examined whether heterogeneity in political
a¢ liation within a peer group had any e¤ect on individual political engagement and political a¢ liation.
It did not as far as political a¢ liation is concerned, but we did …nd, contrary to Mutz (2002a), that
heterogeneity appeared to discourage casting an invalid vote (the results are reported in Table A5 in
the electronic Appendix).
19
with those of their peers.
An individual’s political identi…cation is associated with the identi…cation of his or
her peer group in our data; and so too is an individual’s political engagement with the
engagement of his or her peers. But for di¤erent reasons we argue that neither association should trouble liberal democracies; or sound a more general warning bell for those
who take individual preferences as given. In the case of the correlation between peer
political identi…cation and individual political identi…cation, we …nd that this disappears once we control for selection biases in the way that groups are constituted. The
more wide ranging association between peer political engagement and both individual
engagement and individual political identi…cation survives these controls for selection
biases. It is real, in this sense. But we have argued it is best explained as a consequence of the transmission of information about candidates between those in a group
who initially know more about the political process and those who know less. As such
these peer in‡uences seem a natural by-product of social interaction rather than an
example of preference conformism within a group and so should not worry supporters
of democracy. Indeed, the reverse is the case because this type of peer in‡uence is likely
to be good for the functioning of democracies. Indeed, the reverse is the case because
this type of peer in‡uence is likely to be good for the functioning of democracies. This
is an especially striking conclusion. Our subjects were young adults, embarking on a
new, important and unfamiliar phase in their lives and these seem to be precisely the
uncertain and portentous circumstances where one might expect individual sensitivity
to the cues of others.
References
[1] Banerjee, Abhijit, Selvan Kumar, Rohini Pande and Felix Su. 2010. "Do Informed
Voters Make Better Choices? Experimental Evidence from Urban India." Working
Paper, MIT.
[2] Beck, Paul A. 2002. “Encouraging Political Defection: The Role of Personal Discussion Networks in Partisan Desertions to the Opposition Party and Perot Votes
in 1992.”Political Behavior, 24(4): 309-337.
[3] Bertrand, Marianne, and Sendhil Mullainathan. 2004. "Are Emily and Greg More
Employable Than Lakisha and Jamal? A Field Experiment on Labor Market
Discrimination." American Economic Review, 94(4): 991–1013.
20
[4] Bond, Robert M., Christopher J. Fariss, Jason J. Jones, Adam D. I. Kramer,
Cameron Marlow, Jaime E. Settle and James H. Fowler. 2012. “61-million-person
experiment in social in‡uence and political mobilization.”Nature 489, 295–298.
[5] Carrell, Scott E., Mark Hoekstra and James E. West. 2011. “Is Poor Fitness Contagious? Evidence from Randomly Assigned Friends.”Journal of Public Economics,
95(7-8): 657-663.
[6] Cho, Wendy, James Gimpel and Joshua Dyck. 2006. “Residential Concentration,
Political Socialization, and Voter Turnout.”Journal of Politics, 68: 156–167.
[7] Cho, Wendy K.T. 2003. “Contagion E¤ects and Ethnic Contribution Networks.”
American Journal of Political Science, 47(2): 368-387.
[8] Da Silveira, Bernardo S. and De Mello, João M. P. “Campaign Advertising and
Election Outcomes: Quasi-Natural Experiment Evidence from Gubernatorial Elections in Brazil.”Review of Economic Studies, 78: 590-612.
[9] Funk, Patricia. 2010. “Social Incentives and Voter Turnout: Evidence from the
Swiss Mail Ballot System.” Journal of European Economic Association, 8 (November): 1077-1103.
[10] Gay, Claudine. 2009. Homo Politicus Is Not An Island, in The Future of Political
Science: 100 Perspectives, edited by Gary, King, Kay Schlozman, and Norman
Nie. New York: Routledge.
[11] Gentzkow, Matthew. 2006. “Television and Voter Turnout.” Quarterly Journal of
Economics 121(3): 931-72.
[12] Gerber, Alan; Huber, Gregory, Doherty, David and Conor Dowling. 2012. “Disagreement and the Avoidance of Political Discussion: Aggregate Relationships
and Di¤erences across Personality Traits.”American Journal of Political Science,
56(4): 849-874.
[13] Gerber, Alan, Donald Green and Christopher Larimer. 2008. "Social Pressure and
Voter Turnout: The Results of a Large Scale Field Experiment." American Political
Science Review, 102 :33-48.
[14] Guryan, Jonathan, Kory Kroft, and Matt Notowidigdo. 2009. “Peer E¤ects in the
Workplace: Evidence from Random Groupings in Professional Golf Tournaments.”
American Economic Journal: Applied Economics, 1(4): 34–68.
21
[15] Hung, Angela A., and Charles R. Plott. 2001. “Information Cascades: Replication and an Extension to Majority Rule and Conformity-Rewarding Institutions."
American Economic Review, 91(5): 1508-1520.
[16] Huckfeldt, Robert, and Jeanette Morehouse Mendez. 2008. “Moths, Flames, and
Political Engagement: Managing Disagreement within Communication Networks.”
Journal of Politics, 70: 83-96.
[17] Huckfeldt, Robert. 2007. “Unanimity, Discord, and the Communication of Public
Opinion.”American Journal of Political Science, 51: 978-995.
[18] Huckfeldt, Robert, Paul Allen Beck, Russell J. Dalton and Je¤rey Levine. 1995.
"Political Environments, Cohesive Social Groups, and the Communication of Public Opinion." American Journal of Political Science, 39 (4): 1025-1054.
[19] Huckfeldt, Robert and John Sprague. 1987. “Networks in Context: The Social Flow
of Political Information.” The American Political Science Review, 81(4): 11971216.
[20] Kenny, Christopher B. 1994.“The Microenvironment of Attitude Change.” The
Journal of Politics, 56 (3): 715-728.
[21] Klofstad, Casey A. 2010. “The Lasting E¤ect of Civic Talk on Civic Participation:
Evidence from a Panel Study.”Social Forces, 88 (5): 2353-2375.
[22] Klofstad, Casey A. 2009. “Civic Talk and Civic Participation: The Moderating
E¤ect of Individual Predispositions.”American Politics Research, 37: 856-878.
[23] Lassen, David. 2005. “The E¤ect of Information on Voter Turnout: Evidence from
a Natural Experiment.”American Journal of Political Science, 49(1): 103-18.
[24] Lyle, David. 2007. "Estimating and Interpreting Peer and Role Model E¤ects from
Randomly Assigned Social Groups at West Point." Review of Economics and Statistics, 89(2): 289-299.
[25] Lyle, David S. 2009. "The E¤ects of Peer Group Heterogeneity on the Production of
Human Capital at West Point." American Economic Journal: Applied Economics,
1(4): 69–84.
[26] MacKuen, Michael and Courtney Brown. 1987. “Political Context and Attitude
Change.”American Political Science Review, 81(2): 471-490.
22
[27] Manski, Charles F. 1993. “Identi…cation of Endogenous Social E¤ects: The Re‡ection Problem.”Review of Economic Studies, 60 (3): 531-542.
[28] Maringoni, Gilberto. "Voto nulo, passividade e conservadorismo." Carta Maior,
August 31, 2010.
[29] Mutz, Diana C. and Je¤ery Mondak. 2006. “The Workplace as a Context for CrossCutting Political Discourse.”The Journal of Politics, 68(1): 140-155.
[30] Mutz, Diana C. 2002a. “The Consequences of Cross-Cutting Networks for Political
Participation.”American Journal of Political Science, 46(4) : 838–855.
[31] Mutz, Diana C. 2002b. "Cross-Cutting Social Networks: Testing Democratic Theory in Practice", American Political Science Review, 96(1) : 111-126.
[32] Nickerson, David W. 2008. “Is Voting Contagious? Evidence from Two Field
Experiments,”American Political Science Review, 102 (1):49-57.
[33] Panagopoulos, Costas. 2010.“A¤ect, Social Pressure and Prosocial Motivation:
Field Experimental Evidence of the Mobilizing E¤ects of Pride, Shame and Publicizing Voting Behavior.”Political Behavior, 32: 369-386.
[34] Sacerdote, Bruce. 2001. “Peer E¤ects with Random Assignment: Result for Darmounth Roommates.”Quarterly Journal of Economics, 116 (2): 681-704.
[35] Salganik, Matthew J. , Peter S. Dodds, and Duncan J. Watts.2006.“Experimental
study of inequality and unpredictability in an arti…cial cultural market.” Science,
311:854-856.
[36] Sinclair, Betsy. 2009. “The Multi-Valued Treatment E¤ects of Political Networks
and Context: When Does a Democrat Vote Like a Republican?." Working Paper,
University of Chicago.
23
Figure 1- Number of stories containing the word “Elections” in the Newspaper O Estado de
São Paulo
Number of news stories containing the word "elections"
160
140
120
100
80
60
40
20
0
30
Table 1 – Summary Statistics Individual and Classroom Level
Mean
Stand Dev
Min
Max
Obs
0.500
0.414
0.486
0.500
5.000
0.420
0.491
0.483
0.107
0.101
0.374
0.498
0
0
0
0
17
0
0
0
0.032
0.056
0
0
1
1
1
1
60
1
1
1
0.614
0.658
1
1
622
622
622
547
626
623
623
623
633
633
622
627
0.489
0.189
33.188
27.282
0.090
0.126
11.164
11.863
0.29
0.03
12
9
0.71
0.50
65
52
48
48
48
39
0.058
3.460
0.308
0.301
0.240
0.364
0.396
0.36
0.36
0.233
1.852
0.161
0.132
0.427
0.481
0.489
0.11
0.100
0
0
0
0
0
0
0
0.03
0.06
1
10
0.71
0.67
1
1
1
0.57
0.66
625
634
486
622
622
622
622
472
472
3.490
2.590
4.620
2.125
2.167
2.152
0
0
0
7
7
7
632
630
631
0.047
0.063
0.083
0.047
0.203
0.066
0.031
0.115
0.342
0.212
0.243
0.277
0.212
0.403
0.249
0.175
0.319
0.475
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
635
635
635
635
635
635
635
635
635
Panel A: Pre-determined characteristics and preferences
Individual Characteristics
Female
White
Mother has a college degree
Have a partisan parent
Age
Right-wing
Center
Left-wing
Right-wing Index
Left-wing Index
Intends to Cast and Invalid Vote
Intends to watch Political Campaign on TV
0.491
0.781
0.620
0.510
20.880
0.228
0.402
0.370
0.376
0.362
0.168
0.451
Panel B: Classroom Composition (Peer variables)
Have a partisan parent
Right-wing oriented
Number of Respondents (in t-1)
Number of Respondents (in t)
Panel C: Outcomes
Cast an Invalid Vote
Asymmetric Mistakes
Mistakes on Own Intended Candidate
Mistakes on Remaining Candidates
Right-wing
Center
Left-wing
Right-wing Index
Left-wing Index
Number of days follows politics on
TV
Newspaper
Internet
College Majors
Architecture
Business Administration
Economics
History
Law
Literature
Mathematics
Physics
Sociology
Notes: 1) The sample refers to students in the panel. 2) The explanation of political knowledge and the right- and left-wing
indexes are in the text.
Table 2 - Tests for Random Assignment of Peers Among Classrooms
Coefficient [Stand Error] on Peer Variable:
Has a partisan parent
Right wing
[1]
[2]
-0.0537
[0.1120]
-0.2507
[0.0964]**
-0.2147
[0.1640]
-0.0266
[0.0394]
-0.0092
[0.0320]
0.1407
[0.8045]
0.155
[0.1242]
0.0239
[0.1090]
-0.3068
[0.1449]**
0.0896
[0.0822]
0.0065
[0.2080]
0.0148
[0.0491]
0.0207
[0.0505]
-0.01
[1.083]
-0.0863
[0.2240]
0.0246
[0.1383]
-0.0219
[0.2023]
-0.1065
[0.1295]
0.9305
[3.0249]
-0.0981
[0.1166]
-0.1816
[0.1785]
-0.1023
[0.1933]
-2.732
[2.7207]
0.1471
[0.1430]
Dependent Variable: Pre-determined individual characteristic and habits
[1]
Has a partisan parent
[2]
Right-wing
[3]
Center-oriented
[4]
Right-wing index
[5]
Left-wing index
[6]
Number of days follows politics on TV
[7]
Intends to watch Political Campaign on TV
[8]
Intends to cast an invalid vote
[9]
Demographics
Female
[10]
Mother has a college degree
[11]
Age
[12]
White
Notes: 1) The Table reports OLS estimates from separate regressions of the relevant pre-determined individual characteristic
on the proportion of classmates that have a partisan parent (Column 1) and the proportion of right-wing classmates (Column 2).
All regressions include major-class fixed effects and average value of the peer characteristics among students in the same
major-class (excluding himself). 2) Standard errors clustered at the classroom level are in brackets; ** 5% significance,
* 10% significance.
Table 3: Impact of Classmates' Ideologies on Students Political Orientations
Selected Controls
Peer Variables
% Right-Wing Classmates
Dependent Variable: Self-declaring Right-wing by Election Time
(1)
(2)
(3)
(4)
(5)
(6)
0.9166
[0.0337]**
0.3322
[0.1131]**
0.065
[0.2365]
0.0007
[0.2637]
% Right-Wing Classmates X Right-wing
% Right-Wing Classmates
with a Partisan Parent
Pre-determined ideologies
Right-wing
-0.0459
[0.2639]
0.1945
[0.3839]
-0.0163
[0.3907]
Right-W Index
(7)
0.0964
[0.0885]
-0.2427
[0.1834]
0.5260
[0.0460]**
-0.1060
[0.0253]**
Left-wing
0.5080
[0.0494]**
-0.1072
[0.0270]**
0.5005
[0.0520]**
-0.0851
[0.0294]**
0.4472
[0.1300]**
-0.0863
[0.0292]**
0.4973
[0.0535]**
-0.0927
[0.0305]**
Center (omitted)
Right-wing Index
0.4195
[0.0604]**
0.0222
[0.0826]
Left-wing Index
Additional Controls
Individual characteristics
% Classmates with a Partisan Parent
Major-Class Fixed effects
Observations
no
no
no
612
yes
no
no
563
yes
no
yes
563
yes
yes
yes
497
yes
yes
yes
497
yes
yes
yes
474
yes
yes
yes
375
Notes: 1) Each column represents the result from a separate OLS regression. 2) Standard errors in brackets are clustered at the classroom level.
3) Right-W Index and Left-W Index are constructed ideological indexes explained in Section 2.3.
4) Individual characteristics include gender, race, age, income, mother education and indicadors for students’ pre-determined inclination - right- and left-wing.
5) ** Statistically significant at 5%; * Statistically significant at 10%
Table 4 - Partisan Parent Peer Effects
Coefficient on the Variable Proportion of Classmates with a Partisan Parent
Sample
Consumption of media
All
Dependent Variables:
[1]
Political Knowledge
Asymmetric Mistakes
[2]
Mistakes on Own Intended Candidate
[3]
Mistakes on Remaining Candidates
[4]
Voting Participation and Ideology
Cast an Invalid Vote
[5]
Center-Oriented
[6]
Left-wing Index
[7]
Right-wing Index
Has a Partisan Parent
Yes
No
Number of days per week in any media outlet (in t-1)
Less than 10 days
More than 10 days
[1]
[2]
[3]
[4]
[5]
-0.5128
[0.5055]
506
0.3919
[0.7010]
402
-1.016
[0.4835]**
500
1.000
[1.2966]
254
0.0762
[0.1316]
210
-0.2379
[0.5880]
253
-2.590
[1.3050]**
252
0.0083
[0.1570]
192
-1.0933
[0.4379]**
247
-2.175
[1.595]
223
0.0705
[0.2419]
161
-0.9955
[0.6552]
220
-0.0322
[1.424]
283
0.0819
[0.1333]
241
-0.7068
[0.4315]
280
-0.2322
[0.1049]**
496
0.4389
[0.1621]**
497
-0.1369
[0.0603]**
376
-0.0784
[0.0583]
376
0.0442
[0.0871]
252
0.2444
[0.2796]
250
-0.1056
[0.1102]
186
-0.1391
[0.1119]
186
-0.466
[0.1710]**
244
0.7689
[0.3293]**
247
-0.1678
[0.0616]**
190
-0.0662
[0.0719]
190
-0.4980
[0.1734]**
216
0.4636
[0.2958]
218
-0.2722
[0.0498]**
169
0.0345
[0.0781]
169
0.0306
[0.1379]
280
0.4157
[0.2311]*
279
-0.1612
[0.0890]*
207
-0.1786
[0.0979]*
207
-1.3711
[1.003]
503
-0.2970
[1.0532]
501
0.5260
[1.537]
503
-3.7228
[1.479]**
252
-0.6117
[2.082]
252
0.2559
[2.3965]
252
0.9683
[1.468]
251
0.4862
[1.3789]
249
0.7812
[1.573]
250
-2.959
[2.1147
283
2.4767
[1.341]*
281
-0.3608
[1.2097]
283
-2.2801
[1.6696]
220
-3.789
[1.4472]**
220
-0.9321
[2.1695]
220
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
Consumption of Media
[8]
Number of days follows politics on TV
[9]
Number of days follows politics on Newspaper
[10]
Number of days follows politics on Internet
Individual Characteristics
% Right-Wing Classmates
Major-Class Fixed effects
Notes: 1) Each entry represents the result from a separate OLS regression. 2) Standard errors in brackets are clustered at the classroom level.
3) Individual characteristics include gender, race, age, income, mother education, indicators for students’
pre-determined political inclination - right- and left-wing and dummies indicating whether the student has a partisan parent.
4) The number of observations in each regression are reported in italics.
5) ** Statistically significant at 5%; * Statistically significant at 10%
Appendix
Figure A1: Presidential Race in Brazil 2010: Opinion Polls – Vote Intention – Nov/2009 to
Sept/2010
Source: Supreme Electoral Court 36 Figures A2 - Actual and Simulated Data – Actual Means and Simulated Confidence Intervals
– Outcome and Predetermined Variables
Pre-determined characteristics
Center-Oriented
.8
.6
.4
.2
0
0
.2
.4
.6
.8
Means and Confidence Interval
1
1
Left-Wing Oriented
Bus1 Arc1 Soc2Soc1Law1Law2Eco1Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Bus1 Arc1 Soc2Soc1Law1Law2Eco1Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Female
.8
.6
.4
.2
0
0
.1
.2
.3
.4
.5
.6
.7
.8
Means and Confidence Interval
.9
1
1
Right-Wing Oriented
Bus1 Arc1 Soc2Soc1Law1Law2Eco1Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Bus1 Arc1 Soc2Soc1Law1Law2Eco1Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Has a Partisan Parent
.8
.6
.4
.2
0
0
.2
.4
.6
.8
Means and Confidence Interval
1
1
Has a Mother With College Degree
Bus1 Arc1 Soc2Soc1Law1Law2Eco1 Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Bus1 Arc1 Soc2Soc1Law1Law2Eco1 Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Outcomes
Cast an Invalid Vote
.8
.7
.6
.5
.4
.3
0
0
.1
.2
Means and Confidence Interval
.1
.2
.3
.4
.5
.6
.7
.8
.9
.9
1
1
Percentage of Correct Answers in the Political Quiz
Bus1 Arc1 Soc2Soc1Law1Law2Eco1 Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Bus1 Arc1 Soc2Soc1Law1Law2Eco1 Eco2Phy1 Phy2 His2 His2 Lit1 Lit2 Mat1
Major-Class
Note:1) Bus, Arc, Soc, Law, Eco, Phy, His, Lit and Mat refers respectively to majors Business Administration,
Architecture, Sociology, Law, Economics, Physics, History, Literature and Math. 1 and 2 refers to day and night. 2)
Figures above show the average characteristics of the respondents in the actual panel by major-class. Means are
positioned with respect to the confidence interval constructed based on the simulated random panel groups as
explained in the text.
38
Table A1 - Means reported in the Sample and in USP Adminsitrative Reports
Major - Class
Business
morning
Architecture
morning
(A) First Survey
(B) Panel
(C) Administrative records
48.2
46.3
41.0
73.0
76.7
70.0
51.9
44.1
54.0
39.3
61.1
38.2
50.0
41.5
46.2
26.7
37.5
36.6
26.3
30.8
26.7
20.0
17.6
15.6
29.7
7.7
23.3
24.3
11.8
17.0
39.1
35.0
46.9
45.3
52.4
35.7
71.7
72.0
68.5
58.1
58.2
52.2
29.7
35.0
38.8
p-value ([B]=[C])
p-value ([A]=[C])
0.50
0.10
0.40
0.50
0.56
0.77
0.48
0.83
0.51
0.31
0.88
0.00
0.59
0.94
0.76
0.32
0.07
0.27
0.52
0.31
0.29
0.21
0.15
0.07
0.35
0.27
0.33
0.07
0.73
0.06
(A) First Survey
(B) Panel
(C) Administrative records
71.4
80.0
64.0
78.4
80.0
61.3
74.5
55.9
54.0
49.5
77.8
47.3
82.9
86.3
67.6
66.3
72.2
55.3
73.7
71.8
55.6
51.8
50.0
50.0
50.0
69.2
41.7
40.5
47.1
35.0
69.2
80.0
53.1
44.2
42.9
39.2
55.1
56.6
42.9
41.6
38.0
31.1
26.4
30.0
34.7
p-value ([B]=[C])
p-value ([A]=[C])
0.02
0.05
0.02
0.00
0.03
0.00
0.33
0.68
0.00
0.00
0.00
0.00
0.03
0.00
1.00
0.75
0.06
0.19
0.35
0.50
0.01
0.01
0.74
0.33
0.00
0.00
0.24
0.00
0.66
0.08
(A) First Survey
(B) Panel
(C) Administrative records
22.9
15.0
24.0
18.6
16.7
20.7
17.6
26.5
28.0
29.7
16.7
26.3
13.7
13.7
17.8
24.7
25.0
22.5
21.1
25.6
17.8
34.1
38.2
33.3
35.9
15.4
35.0
32.4
29.4
30.0
20.0
5.0
21.6
34.7
19.0
32.1
29.3
27.6
27.2
34.2
42.3
34.4
40.7
35.0
40.7
p-value ([B]=[C])
p-value ([A]=[C])
0.12
0.75
0.56
0.59
0.23
0.06
0.98
0.49
0.41
0.12
0.63
0.51
0.28
0.49
0.56
0.87
0.08
0.88
0.96
0.76
0.00
0.75
0.15
0.59
0.92
0.46
0.19
0.95
0.61
0.99
(A) First Survey
(B) Panel
(C) Administrative records
5.9
2.4
8.0
6.1
6.9
16.7
8.2
12.1
20.0
20.5
13.3
19.0
6.5
4.1
8.8
12.3
9.0
12.0
12.5
8.1
16.6
13.1
18.2
19.9
30.0
33.3
31.8
32.4
23.5
46.0
19.0
21.1
24.8
37.9
42.9
39.3
27.8
28.3
41.4
27.4
28.4
44.8
41.8
55.0
57.1
p-value ([B]=[C])
p-value ([A]=[C])
0.03
0.30
0.05
0.00
0.48
0.00
0.24
0.74
0.11
0.22
0.39
0.92
0.07
0.30
0.80
0.07
0.92
0.76
0.05
0.09
0.70
0.25
0.75
0.78
0.00
0.00
0.00
0.00
0.86
0.00
(A) First Survey
(B) Panel
(C) Administrative records
23.5
26.8
31.0
31.3
24.1
28.7
28.6
27.3
28.0
22.7
40.0
30.9
18.8
20.4
26.3
21.3
17.9
25.5
20.8
24.3
26.6
27.4
27.3
23.3
28.3
41.7
30.0
48.6
64.7
30.0
34.9
36.8
38.4
28.4
28.6
30.0
37.8
42.0
26.6
38.1
41.8
30.9
35.2
30.0
24.5
p-value ([B]=[C])
p-value ([A]=[C])
0.55
0.04
0.58
0.58
0.37
0.93
0.65
0.07
0.32
0.01
0.11
0.20
0.75
0.24
0.62
0.41
0.45
0.78
0.01
0.03
0.89
0.57
0.89
0.74
0.00
0.00
0.08
0.03
0.61
0.04
(A) First Survey
(B) Panel
(C) Administrative records
29.4
29.3
32.0
35.4
41.4
31.3
34.7
24.2
27.0
28.4
40.0
23.6
26.5
28.6
28.9
31.6
34.3
30.6
31.9
37.8
25.5
31.0
27.3
35.6
35.0
25.0
30.0
18.9
11.8
16.0
25.4
26.3
24.7
20.0
14.3
18.6
24.5
20.3
19.7
24.3
19.4
19.2
18.7
10.0
12.3
p-value ([B]=[C])
p-value ([A]=[C])
0.71
0.51
0.29
0.40
0.34
0.27
0.93
0.32
0.96
0.48
0.53
0.79
0.14
0.25
0.30
0.36
0.71
0.42
0.61
0.66
0.88
0.90
0.59
0.74
0.86
0.09
0.97
0.07
0.74
0.12
(A) First Survey
(B) Panel
(C) Administrative records
41.2
41.5
29.0
27.3
27.6
23.3
28.6
36.4
25.0
28.4
6.7
26.5
48.2
46.9
36.0
34.8
38.8
31.9
34.7
29.7
31.3
28.6
27.3
21.2
6.7
0.0
0.1
0.0
0.0
0.1
20.6
15.8
12.1
13.7
14.3
12.1
10.0
9.4
12.3
10.2
10.4
5.1
4.4
5.0
6.1
p-value ([B]=[C])
p-value ([A]=[C])
0.12
0.00
0.62
0.38
0.02
0.59
0.25
0.69
0.14
0.00
0.25
0.45
0.84
0.55
0.45
0.14
--0.64
-----
0.67
0.10
0.78
0.66
0.25
0.23
0.16
0.01
0.83
0.43
Characteristcs
Female
Sociology
morning
evening
Law
morning
evening
Economics
morning
evening
Physics
morning
evening
History
morning
evening
Literature
morning
evening
Mathematics
morning
Mother has college education
Mother has high school education
Income (up to 5 min. wage)
Income (5 up to 10 min. wage)
Income (10 up to 20 min. wage)
Income (above 20 min. wage)
Table A2 – Students’ Pre-determined Characteristics According to Parents’ Preferences
Number of
Has a
Does Not Have a
Observations
Partisan Parent
Partisan Parent
[1]
[2]
[3]
[4]
[ 3 ]- [ 4 ]
0.464
[0.498]
0.570
[0.495]
21.840
[6.40]
0.763
[0.425]
1,556
0.456
[0.498]
0.629
[0.483]
21.330
[5.56]
0.771
[0.420]
0.458
[0.498]
0.521
[0.499]
22.320
[7.02]
0.769
[0.421]
-0.002
0.229
[0.420]
0.283
[0.451]
0.486
[0.500]
0.495
[0.500]
0.144
[0.351]
0.187
[0.390]
0.469
[0.499]
0.343
[0.475]
0.387
[0.487]
0.187
[0.390]
All
Demographics
Female
Mother has a college degree
Age
White
Political Preferences and Habits
Has a partisan parent
Right-wing oriented
Center-oriented
Left-wing oriented
Intends to watch Political Campaign on TV
Intends to Cast and Invalid Vote
0.491
[0.500]
0.206
[0.404]
0.386
[0.487]
0.407
[0.491]
0.426
[0.494]
0.183
[0.387]
1,565
1,548
1,545
0.108
**
-0.990 **
0.002
1,374
1,557
1,557
1,557
1,570
1,549
Notes: 1) Standard errors in brackets are clustered at the classroom level. 2) The statistics refer to information provided in the first survey.
3) ** Statistically significant at 5%; * Statistically significant at 10%
0.042
*
-0.186 **
0.143
**
0.108
**
-0.043 **
Major-period
Number of
Classrooms
Architecture - day
Business - day
Economics - day
Economics - night
History - day
History - night
Law - day
Law - night
Literature - day
Literature - night
Mathematics - night
Physics - day
Physics - night
Sociology - day
Sociology - night
All Classrooms
4
4
2
2
2
3
3
4
6
6
2
2
2
2
4
48
Appendix A3- Descriptives
Average Number of
Variance by Peer Variable
Respondents
% Right-wing % has a partisan
per Classroom
parent
26.5 [3.52]
35.85 [6.57]
37.5 [0.50]
42.5 [0.50]
32.5 [0.50]
35.47 [10.36]
45.52 [7.23]
49.15 [15.26]
38.88 [8.18]
31.23 [7.57]
46.97 [4.49]
37.4 [11.91]
21.57 [6.59]
26.29 [4.87]
23.53 [4.23]
37.02 [11.13]
0.1413
0.2436
0.2447
0.2413
0.1528
0.0794
0.2311
0.21
0.1296
0.0803
0.1391
0.1615
0.1576
0.0765
0.0919
0.1716
0.2436
0.2501
0.2382
0.2533
0.2533
0.254
0.2483
0.247
0.2459
0.2509
0.2529
0.2543
0.254
0.2414
0.2532
0.2501
Allocation
Rule
Random algorithm
Alphabetical order
Alphabetical order
Alphabetical order
Random algorithm
Random algorithm
Alphabetical order
Alphabetical order
Random algorithm
Random algorithm
Alphabetical order
Random algorithm
Random algorithm
Alphabetical order
Alphabetical order
Appendix A4 -Effects of Heterogeneous Groups
Political Knowledge
Asymmetric Mistakes
Mistakes on Own Intended Candidate
Mistakes on Remaining Candidates
Voting Participation and Ideology
Cast an Invalid Vote
Center-Oriented
Left-wing Index
Right-wing Index
Individual Characteristics
% Right-Wing Classmates
% Partisan parent Peers
Major-Class Fixed effects
(1)
(2)
(3)
(4)
-0,089
(0,551)
506
-0,027
(0,039)
402
0,013
(0,029)
500
-0,031
(0,552)
506
-0,028
(0,039)
402
0,018
(0,030)
500
-0,160
(0,545)
506
-0,029
(0,039)
402
0,011
(0,029)
500
-0,160
(0.542)
506
-0,028
0,039
402
0,011
(0.029)
500
-0,169
(0,052)**
496
-0,043
(0,142)
497
0,015
(0,036)
376
0,062
(0,043)
376
yes
no
no
yes
-0,164
(0,052)**
496
-0,057
(0,142)
497
0,021
(0,037)
376
0,067
(0,043)
376
yes
no
yes
yes
-0,168
(0,051)**
496
-0,046
(0,143)
497
0,014
(0,036)
376
0,057
(0,043)
376
yes
yes
no
yes
-0,168
(0,052)**
496
-0,046
(0,142)
497
0,014
(0,036)
376
0,057
(0,043)
376
yes
yes
yes
yes
Notes: 1) Each entry represents the result from a separate OLS regression. 2) Standard errors in brackets are clustered at the classroom level.
3) Individual characteristics include gender, race, age, income, mother education, indicators for students’
pre-determined political inclination - right- and left-wing and dummies indicating whether the student has a partisan parent.
4) The number of observations in each regression are reported in italics.
5) The peer variable of cross-cutting information is constructed as the percentage of individuals in the classroom, minus the own subject, that have
a different political view from the subject.